Connect to Impala

​

Use Impala ODBC connector for Cloudera Enterprise with pyodbc

​

Environment setup

1. Visit the Cloudera downloads page to download the Impala ODBC Connector for Cloudera Enterprise to your local machine. For default Domino images of Ubuntu 16.04, download the 64-bit Debian package. Keep track of where you save this file, as you need it in a later step.
2. Create a new public project in your Domino instance to host the driver files for use in Domino environments.
3. In the new project, click browse for files and select the driver file you downloaded earlier to queue it for upload. Click Upload to add it to the project.
4. After the driver file has been added to your project files, click the gear next to it in the files list, then right-click Download and click Copy link address. Save this address as you need it when setting up your environment.
5. Add the following Dockerfile instructions to install the driver and pyodbc in your environment, pasting the URL you copied earlier where indicated on line 5.
   USER root
   
   # download the driver from your project
   RUN mkdir /ref_files
   RUN
   cd /ref_files &&
   wget --no-check-certificate [paste-download-url-from-previous-step-here] &&
   gzip -d clouderaimpalaodbc_2.6.0.1000-2_amd64.deb.gz
   
   # install the driver
   RUN gdebi /ref_files/clouderaimpalaodbc_2.6.0.1000-2_amd64.deb --n
   
   # update odbc.ini file for impala driver
   RUN
   echo "
   
   [Cloudera ODBC Driver for Impala]
   
   Driver=/opt/cloudera/impalaodbc/lib/64/libclouderaimpalaodbc64.so
   
   KrbFQDN=_HOST
   
   KrbServiceName=impala
   " >> /etc/odbcinst.ini
   
   # set up impala libraries
   RUN export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/opt/cloudera/impalaodbc/lib/64
   RUN ldd /opt/cloudera/impalaodbc/lib/64/libclouderaimpalaodbc64.so
   
   # install pyodbc
   RUN pip install pyodbc
   
   USER ubuntu
   

​

Credential setup

• IMPALA_HOST

• IMPALA_PORT

• IMPALA_KERB_HOST

• IMPALA_KERB_REALM

​

Usage

import pyodbc
import os

# fetch values from environment variables
hostname = os.environ['IMPALA_HOST']
service_port = os.environ['IMPALA_PORT']
kerb_host = os.environ['IMPALA_KERB_HOST']
kerb_realm = os.environ['IMPALA_KERB_REALM']

# create connection object
conn = pyodbc.connect('Host=hostname;'
                     +'DRIVER={Cloudera ODBC Driver for Impala};'
                     +'PORT=service_port;'
                     +'KrbRealm=kerb_realm;'
                     +'KrbFQDN=kerb_host;'
                     +'KrbServiceName=impala;'
                     +'AUTHMECH=1',autocommit=True)

# if you see:
# 'Error! Filename not specified'
# while querying Impala using the connection object,
# add the following configuration line:
#
# conn.setencoding(encoding='utf-8', ctype=pyodbc.SQL_CHAR)

# if your Impala uses SSL, add SSL=1 to the connection string
# conn = pyodbc.connect('Host=hostname;'
#                      +'DRIVER={Cloudera ODBC Driver for Impala};'
#                      +'PORT=service_port;'
#                      +'KrbRealm=kerb_realm;'
#                      +'KrbFQDN=kerb_host;'
#                      +'KrbServiceName=impala;'
#                      +'AUTHMECH=1;'
#                      +'SSL=1;'
#                      +'AllowSelfSignedServerCert=1', autocommit=True)

​

Next steps

• Access external data directly: the general procedure, including where to store credentials.
• Compare direct connections with Domino Data Source connectors for supported data stores.